RESEARCH BRIEF * NOVEMBER 2020

Extending Unemployment Insurance Benefits to Workers in Precarious and Nonstandard Arrangements Katharine G. Abraham Professor of Economics and Survey Methodology, University of Maryland Co-Director, Labor Statistics Program, IZA Research Associate, NBER Susan N. Houseman Vice President and Director of Research, W.E. Upjohn Institute for Employment Research Co-Director, Labor Statistics Program, IZA Member, Task Force on the Work of the Future Research Advisory Board Christopher J. O’Leary Senior Economist, W.E. Upjohn Institute for Employment Research

1

Extending Unemployment Insurance Benefits to Workers in Precarious and Nonstandard Arrangements

Katharine G. Abraham Professor of Economics and Survey Methodology, University of Maryland Co-Director, Labor Statistics Program, IZA Research Associate, NBER Susan N. Houseman Vice President and Director of Research, W.E. Upjohn Institute for Employment Research Co-Director, Labor Statistics Program, IZA Member, Task Force on the Work of the Future Research Advisory Board Christopher J. O’Leary Senior Economist, W.E. Upjohn Institute for Employment Research

Introduction

Many fear that rapid technical change is raising the rate at which workers are displaced from their jobs, as

machines take over many tasks once performed by humans (Brynjolfsson and McAfee, 2014). By shifting

the locus of production of many goods from advanced economies to emerging and developing ones,

globalization also has contributed to worker displacement (see, e.g., Autor, Dorn and Hanson, 2013;

Acemoglu et al., 2016; Pierce and Schott, 2016). Even aside from changes driven by technology and

trade, employers’ increasing reliance on contract workers and on-demand scheduling rather than on

permanent employees who work predictable schedules has added to the precariousness of many workers’

jobs (Weil, 2014; Lambert, Fugiel, and J.R. Henly, 2014; Lambert, Henly, and Kim, 2019).

The current global pandemic threatens to magnify these forces. The adoption of new technologies that

minimize workers’ exposure to the virus may accelerate the process of labor-displacing technological

change. The widespread shift to remote work during the pandemic may lead to lasting changes as well.

Although remote work has some drawbacks, it allows workers to avoid time-consuming and costly

2

commutes, and businesses can realize substantial savings on office space and business travel, among other

benefits. Even after the pandemic is over, some of the shift to remote work seems likely to persist,

contributing to reductions in employment in industries such as hotels, restaurants, and building

maintenance—industries that largely serve other businesses and in which employment is dominated by

low-wage workers (Autor and Reynolds, 2020). New jobs will be created, but many likely will be in

different locations and require different skills from those that have disappeared. Further, as employers

confront a less settled economic environment, the shift toward more precarious employment arrangements

may accelerate, adding to workers’ uncertainties.

These developments underscore the importance of a robust social safety net for those who lose their jobs.

They also point to a greater need for workforce program services that help workers find and possibly

retrain for other jobs. Unemployment insurance (UI) is the principal mechanism for assisting job losers in

the United States.

However, serious weaknesses in the unemployment insurance system have developed, particularly in

recent years. In many states, the average share of lost wages replaced by unemployment insurance

benefits and the maximum length of time the unemployed can receive benefits have both fallen. Some

states have made the process of applying for benefits more difficult. Moreover, large numbers of workers

do not qualify for UI benefits at all. Following the Great Recession of 2008–2009, the UI recipiency

rate—the share of the unemployed receiving unemployment benefits—fell to historic lows. The current

pandemic, which triggered the deepest recession since the Great Depression of the 1930s, has shined a

bright light on the UI system’s inadequacies and the need for reform.

Comparing the U.S. UI system with those of other advanced economies also reveals its inadequacies.

Many other developed countries provide both unemployment benefits that are based on a worker’s prior

earnings and means-tested unemployment assistance once unemployment benefits have been exhausted.

The Organization for Economic Cooperation and Development (OECD) produces regular estimates of the

effective share of unemployed individuals’ lost net earnings that are replaced by unemployment insurance

and other social assistance benefits (the net replacement rate). For a single worker with no children paid

two-thirds of the average wage at the beginning of an unemployment spell, among the 40 countries

ranked, the estimates for 2019 show the United States tied for 29th with Estonia (OECD, 2020).1 Other

OECD research placed the United States 33rd out of 34 OECD countries as of 2014 in the maximum

duration of unemployment insurance benefits, ahead only of Hungary. A ranking that same year of 33

countries by pseudo-recipiency rates, calculated as the ratio of the number of people receiving UI benefits

3

to the number of people unemployed, placed the United States 26th, ahead only of Turkey, the Slovak

Republic, Poland, Hungary, Japan, Slovenia, and Lithuania (OECD, 2018a).

The United States can and should provide greater support to unemployed individuals who are looking for

new jobs or seeking to upgrade their skills to transition to a new type of work. In this research brief, we

provide background on the unemployment insurance system and explain how business practices such as

the use of scheduling algorithms (associated with variable work hours), temporary agency workers,

independent contractors, and online platforms (employing “gig workers”) not only make work more

precarious but also leave many without access to UI benefits. We close with a discussion of ways to

reform the UI system to expand coverage to more workers in precarious and nonstandard arrangements

and of the use of UI benefits as a stipend for those who have lost their jobs and need retraining.

Background on Unemployment Insurance in the United States

The Social Security Act of 1935 established the foundation for the unemployment insurance system in the

United States, which is operated as a federal-state partnership. The five main goals of the UI system are to

(1) provide temporary partial wage replacement during involuntary unemployment, (2) prevent dispersal

of employers’ workforces during temporary layoffs, (3) promote rapid return to work, (4) limit business

downturns by maintaining aggregate purchasing power, and (5) encourage stabilization of employment in

enterprises (O’Leary and Straits, 2004).

The unemployment insurance system is financed through state and federal payroll taxes. States levy what

is called an experience-rated payroll tax on employers. Just as an individual’s car insurance premium is

related to past claims, an employer’s state UI tax rate is based on the level of benefits paid out in the past

to those who have been employees of the organization. The revenues from the state taxes may be used

only to pay UI benefits. The federal tax revenues pay for the administration of the state programs, the

public employment service, and loans made to states when the reserves accrued from UI taxes prove

inadequate to pay benefits. Federal UI tax revenues also may be used to cover extended benefits as

provided by the federal government during periods of high unemployment. Unemployment insurance

programs operate in all states, the District of Columbia, Puerto Rico, and the Virgin Islands.

Federal law imposes minimal requirements on state programs, allowing states wide latitude in their

practices. The result has been a system whose parameters vary greatly across states. At the start of 2020,

the national average weekly benefit amount was $373. Maximum state weekly benefits, however, ranged

4

from a low of $235 in Mississippi to a high of $1,234 in Massachusetts. The maximum potential duration

for which an unemployed person could receive regular UI benefits under state program rules ranged from

just 12 weeks in Florida and North Carolina to 28 weeks in Montana (U.S. Department of Labor, 2020a).

In the earliest years of the program, benefit levels and durations were modest. By the 1950s, enabled by

accumulating system reserves, all states set UI benefits to replace approximately 50 percent of an

unemployed worker’s prior wage for at least 26 weeks. These standards were endorsed by blue-ribbon

advisory councils and supported by employment research (NCUC, 1980; USACUC, 1996). Following

research on work disincentives by economist Martin Feldstein (1974) and his advocacy for change while

on the Council of Economic Advisers in the 1980s, UI benefits became taxable income in 1982, reducing

effective income replacement. Deteriorating public finances, resulting in part from anti-tax fervor in many

states, subsequently led to stagnating nominal benefit levels and wage replacement rates that have fallen

since the early 1980s (O’Leary and Wandner, 2020).

The Great Recession depleted the UI trust funds in many states, which then had to borrow from the

federal government to fulfill their obligations. Following the recession, rather than fully replenishing the

trust funds, 10 states sought to mitigate the impact of the recession on employer taxes by reducing the

maximum duration that the unemployed could potentially receive benefits from the standard 26 weeks

(O’Leary and Kline, 2020).2

Qualifying for Unemployment Insurance

With various expansions since its beginnings in the 1930s, the U.S. unemployment insurance system

today covers almost all organizations with paid employees. The precise laws governing UI coverage vary

from state to state, but there are four main coverage exclusions—agricultural workers on small farms,

household workers with low earnings, certain employees of religious organizations, and the self-

employed. The exclusion of agricultural and household workers from coverage under the legislation that

established the unemployment insurance system was an ugly outgrowth of Southern legislators’

opposition to having the program cover black workers. At the time, such a large share of the black

population worked in agriculture or domestic service that omitting those occupations from coverage

provided a way to deny benefits to black workers in a statute that superficially was racially neutral (Perea,

2011).3

5

The self-employed, who comprise the largest category of workers not covered by the unemployment

insurance system, are a diverse group. These workers include traditional business owners who may have

employees and significant capital investment, independent contractors, gig workers who obtain work

through online platforms or mobile apps, and informal workers who typically perform services for

households, such as maintenance, child- and elder-care, and cleaning. Additionally, even employees who

work for a covered employer will not qualify for unemployment benefits if they do not have sufficient

prior earnings when they are laid off or if they fail to meet other UI eligibility requirements.

In contrast to the United States, a number of European countries have extended unemployment insurance

coverage to the self-employed (Maquet, Maestri, and Thévenot, 2016). In some countries, the self-

employed are covered on the same basis as other workers. In other cases, they must meet more stringent

requirements to quality for unemployment benefits, must make voluntary contributions to be covered, or

are eligible only for means-tested unemployment assistance. Where participation in the unemployment

insurance scheme is voluntary, the program tends to attract the self-employed most likely to need the

protection it offers, raising the cost of participation and leading to low participation rates (OECD, 2018b).

Across the European Union as a whole, the risk of not being eligible for benefits is notably higher for the

self-employed than for permanent, full-time, wage and salary workers. Other groups at high risk of being

ineligible for benefits under the rules of the European systems are those working part-time or on

temporary contracts (Matsaganis et al., 2015).

During the current pandemic, for the first time in the history of the UI system, unemployment insurance

was extended nationally to wage and salary workers who normally would not qualify for benefits and to

the self-employed through a special Pandemic Unemployment Assistance (PUA) program. Problems with

the reporting of data about this new program make it difficult to determine exactly how many people have

been collecting PUA benefits, but a rough estimate is that, at of the end of August 2020, the number was

in the neighborhood of 10 million people (Casselman et al., 2020). This would represent close to 40

percent of the total number of people receiving unemployment benefits as of that date.4 Whatever the

exact figure, the size of the groups normally excluded from benefits is large. We discuss employees who

do not qualify for UI benefits and the self-employed in turn.

Wage and Salary Workers: Determining Eligibility

Workers who lose their jobs with a covered employer must meet certain criteria to be eligible to receive

unemployment benefits. Unemployment insurance eligibility rules are set to ensure that those

6

compensated are strongly attached to the labor force and temporarily jobless through no fault of their

own. They must have been involuntarily separated from their job; with few exceptions, if they quit or

were fired for cause, they do not qualify. Additionally, they must be available for and actively seeking

work and may not refuse an offer of suitable work.

Labor force attachment is demonstrated by meeting what are referred to as monetary eligibility

conditions. Monetary eligibility for unemployment insurance is determined primarily by earnings in a

defined base period. In most states, the standard UI base period is the first four of the previous five

completed calendar quarters before the date of claim for benefits. According to the most recently

available full summary of state UI provisions, as of the beginning of 2019, workers in 37 states and the

District of Columbia who did not qualify by virtue of their earnings in the standard base period were

allowed to have their eligibility evaluated based on their earnings during a more recent period, typically

the most recent four completed calendar quarters. The minimum base period earnings required ranged

from a low of $130 in Hawaii to a high of $6,435 in Arizona to qualify for the minimum UI weekly

benefit amounts of $5 and $172, respectively. States often also impose additional requirements, such as

that the claimant must have earned more than some minimum amount in the quarter of the base period

with the highest earnings, or that the claimant must have had earnings in at least two base-period quarters

(U.S. Department of Labor, 2019).

Nonmonetary requirements ensure that a UI recipient is (1) involuntarily unemployed (i.e., was laid off

from work) or voluntarily left work for good cause, (2) available for work, and (3) actively seeking work.

Along with the monetary eligibility requirements, these conditions determine whether an unemployed

worker initially qualifies for benefits. The second and third conditions must be satisfied on a continuing

basis throughout a period of unemployment. If they are not satisfied in any given week, the worker is

ineligible to receive benefits for that week. As noted, the experience rating of UI tax rates provides an

inducement for employers to get involved in the benefit eligibility process. Employers are notified of UI

claims by all recent employees and have an opportunity to object to their eligibility and provide evidence

that unemployment payments are not due.

Even workers who are in covered employment and satisfy the job separation conditions may have

insufficient base-period earnings to qualify for benefits if they are low-wage or part-time workers. As of

the beginning of 2019, a worker in Arizona, Indiana, Michigan, Ohio, or South Carolina who worked 20

hours per week at the state’s minimum wage for a full half-year would have had insufficient base-period

earnings to qualify for UI benefits. In an additional 23 states, working 20 hours per week for three months

7

at the state’s minimum wage would not have generated sufficient base-period earnings to qualify, even if

other monetary eligibility conditions were satisfied.5

A periodic supplement to the Current Population Survey (CPS), most recently administered in 2018, has

examined who among those reporting being unemployed applies for and receives benefits and, in recent

waves of the supplement, the reasons for not applying for UI benefits.6 In 2018, among those who were

unemployed because they lost a job (including those laid off and those in temporary jobs that ended),

slightly over half reported applying for benefits, and about 70 percent of those reported receiving UI

benefits. Among the unemployed who had lost a job but did not apply for benefits, slightly more than half

indicated that they believed, for various reasons, they would not qualify. The two most common reasons

cited were that their work history was insufficient to qualify for benefits (19.1 percent) and that the nature

of their job separation (e.g., they quit or were fired) disqualified them for benefits (18.4 percent).7

In addition to having a sufficient work history, to qualify for unemployment benefits workers in many

states must not only be available to work, they must be available for full-time work. A worker who loses a

part-time job and is seeking a new part-time job may not qualify for UI benefits if the state requires

availability for full-time work. This can be the case even if the person’s availability is limited because of

family obligations such as child- or elder-care.

The American Recovery and Reinvestment Act of 2009 offered cash incentives to states for UI

modernization changes that broadened eligibility (O’Leary, 2011). One modernization option that helped

states qualify for an incentive payment was to relax the availability requirement to allow unemployed

persons who normally worked part-time to restrict availability to other part-time jobs. Twenty-five states

qualified for modernization payments by permitting part-time availability. In some cases, that means

workers whose hours are cut or who lose a second part-time job may qualify for partial unemployment

compensation. The unemployment insurance programs in every state allow for such benefits, but as of the

beginning of 2019, workers in 25 states who earned more than the state’s statutory weekly benefit amount

would have received no partial unemployment compensation. The cutoffs in most other states were only a

little higher (West et al., 2016; U.S. Department of Labor, 2019). Weekly benefit amounts typically are no

more than half of weekly earnings and less for those with higher earnings. Limiting eligibility for partial

benefits in this way means, for example, that a person who holds two, equal-paying, 20-hour-a-week jobs

and loses one of them would not receive any partial unemployment payment. The same generally would

be true for a worker whose hours on a single primary job were cut in half.8

8

The Self-Employed

The self-employed include anyone who is not a W-2 employee (among other things, a W-2 employee is

someone whose employer must deduct certain taxes from his or her pay). Traditionally, the main

justification for excluding the self-employed from UI coverage has been the so-called problem of moral

hazard. The self-employed generally have more discretion over their work hours than W-2 employees,

and so in this case the person at risk of unemployment also controls exposure to the risk. A competitive

unemployment insurance market for the self-employed arguably would collapse because of adverse

selection; only people who intended to stop working would participate. Even a public UI program with

compulsory participation could be expensive to finance because of moral hazard.

Despite these concerns, many European countries have chosen to extend unemployment insurance

coverage to the self-employed. The moral hazard problems just noted are dealt with in part by requiring

the self-employed to register with the employment service and provide evidence of active job search. In

the United States, prior to the current pandemic, unemployment insurance had only ever been extended

broadly to the self-employed as a temporary measure in areas affected by natural disasters, through the

federal Disaster Unemployment Assistance program. In these events, unemployment was so widespread

that the moral hazard problem was deemed minimal.

Distinguishing who is an employee and who is self-employed as an independent contractor is not always

a straightforward task. Generally, classification depends on who controls work activities and the

organization’s finances, the nature of the decision-making relationship between the parties, and workers’

opportunity for profit or loss. There is, however, a continuum of work arrangements along these various

dimensions. At one end of the spectrum are employees of an organization that directs all of their work

activities, covers all of their business-related expenses, and controls revenues. At the other end are

business owners and some independent contractors who solicit and sell products and services to clients,

determine their work activities, and enjoy the profits or bear the losses of their business. In between these

extremes are independent contractors who sometimes depend heavily on just one or a small number of

clients for work and have only partial discretion over their work activities and schedules. In these cases,

the line between W-2 and self-employment can blur. 9

Complicating matters, the test used to determine independent contractor status may vary for different

purposes and across states. For instance, somewhat different criteria may be used to determine whether a

worker is an employee for social security and Medicare tax purposes than are used for workers’

9

compensation purposes or for determining coverage under the Fair Labor Standards Act, which provides

wage and hours protections for employees, among other things. And states use several different tests to

determine employee status for the purpose of unemployment insurance (de Silva et al., 2000). Although

rare, it is thus possible for a worker to be classified as an independent contractor for some purposes and as

an employee for others.

Much of the current policy debate over worker classification has focused on platform companies, whose

workers often occupy a gray zone, displaying characteristics of both employees and the self-employed.

California’s Assembly Bill 5 (AB5), signed into law in September 2019, clarified when firms must

classify workers as employees, and proponents have argued it will reduce misclassification. In most cases,

the law required platform companies to reclassify their independent contractors as employees. A

successful initiative on the ballot in California in November 2020 (funded by Doordash, Uber, and Lyft),

however, exempts app-based drivers from AB5. Instead, the initiative’s language stipulates that these

drivers will continue to be classified as independent contractors and be covered by labor and wage

policies specific to rideshare drivers. In September 2020, in an apparent response to AB5 and pressure

from platform companies (Rosenberg, 2020), the U.S. Department of Labor issued a proposed

interpretative rule that similarly would make it easier to categorize certain workers, such as gig workers,

as independent contractors (Penn, 2020).

Despite the recent policy attention given to the classification of platform workers, this group represents a

very small share of the workforce.10 An issue of arguably greater policy relevance involves not the legal

test used to classify workers but the weak enforcement of worker classification laws. The likelihood that

an employer’s classification of its workers will be challenged is low, and there are concerns that many

firms misclassify employees as independent contractors out of ignorance or as part of a deliberate

business strategy (see, e.g., General Accountability Office, 2009; Carré, 2015; National Employment Law

Project, 2017). The magnitude of independent contractor misclassification is not easily documented.

Many of the available estimates are derived from state unemployment insurance agency audits. Studies

based on UI audits in 21 states found that, as of 2012, between 11 percent and 30 percent of employers

misclassified workers as independent contractors (Carré, 2015, 9).11 Moreover, the prevalence of

misclassification is high among these employers. A study of 2003–2004 Michigan audit data found that

30 percent of audited employers misclassified at least one employee as an independent contractor; among

those employers, 24 percent of employees were misclassified (Belman and Block, 2009).

Misclassification appears to be particularly widespread in some industries such as construction,

transportation, and home services. Misclassification is also common when long subcontracting chains

10

separate the client from the workers who perform services for the client (Weil, 2014, 2017; Bernhardt et

al., 2009). Requiring that workers being treated as independent contractors form limited liability

companies of questionable validity is a common tactic used to help shield companies from worker

misclassification lawsuits (Weil, 2017; Armstrong, Elliott, and Tobin, 2020).

These considerations make it important to know the number of people working under self-employment

arrangements, particularly independent contractor or informal, nonemployee arrangements. The

Contingent Worker Supplement (CWS) to the CPS is the main source of information on independent

contractors in the United States. Administered six times between 1995 and 2017, the CWS estimates that

the share of the workforce in independent contractor arrangements ranged from 6.3 percent to 7.4 percent

during this period and showed no consistent trend. According to CWS data, the share of workers who are

self-employed but not independent contractors has declined over time from 6.2 percent in 1995 to 4.4

percent in 2017 (Abraham and Houseman, forthcoming). Other data sources suggest, however, that the

CPS and CWS may understate the size of the independent contractor and other self-employed workforces

and may have missed some growth in the former. Research using administrative tax data shows higher

levels of self-employment and growth in the share of workers with self-employment income (Abraham et

al., in press; Abraham et al., forthcoming; Jackson, Looney, and Ramnath, 2017; Katz and Krueger, 2019;

Lim et al., 2019). The growth appears to arise primarily from those who are in independent contractor

arrangements (Jackson, Looney, and Ramnath, 2017; Lim et al., 2019).

One potential reason for these discrepancies is that the CPS may misidentify many individuals who are

classified as independent contractors (whether legally classified or not) as employees. In the CPS, those

who report being employed are asked “Were you employed by government, by a private company, a

nonprofit organization, or were you self-employed or [if applicable] working in the family business?”

Those working as independent contractors may not think of themselves as being self-employed, a term

some may associate with business ownership. Moreover, it would be accurate for those working on a

contract basis for an organization to respond that they are employed by the organization; if they do, they

will be coded as wage and salary workers because it is implicitly assumed that an individual working for

an organization is an employee. Evidence from a Gallup survey module on contract work suggests the

incidence of such miscoding is substantial (Abraham, Hershbein, and Houseman, 2019).

11

Reasons for the Low and Declining Share of the Unemployed Who Receive Unemployment Insurance Benefits

An oft-cited indicator of problems with the U.S. unemployment insurance system is the low and declining

share of the unemployed who receive UI benefits. Figure 1 displays the average annual share of

unemployed individuals receiving benefits under regular state UI programs over the last four decades.12

Because only unemployed persons who have lost a job are eligible for unemployment insurance benefits,

we would expect the recipiency rate to be well below 100 percent. That said, while the recipiency rate is

sensitive to the stage of the business cycle—rising somewhat during recessions when a higher share of the

unemployed are job losers—its long-term trend has been downward. The decline was especially large

following the Great Recession, and since 2011, the UI recipiency rate has been consistently below 30

percent. As previously noted, a higher share of the unemployed receive benefits in almost every other

developed country than in the United States.

A variety of reasons have been offered for the recent decline in the UI recipiency rate. Figure 2, which

displays the 2018 recipiency rate by state, suggests that the low aggregate rate can be traced, at least in

part, to policy decisions in some states. In a year when all states’ economies were relatively strong, the UI

recipiency rate exceeded 40 percent in nine states, but fell to between just 10 percent and 15 percent in

eight states, most located in the southeast. As noted, the high level of unemployment in the Great

Recession depleted state UI trust funds. Some states responded by reducing the generosity of benefits or

the maximum duration they can be received. Historically, the stated goal of the UI system was to set

benefit levels at approximately 50 percent of lost wages for most UI recipients (Blaustein, 1993). Recent

research finds that, balancing the value to workers of being able to better maintain consumption while

unemployed against the disincentive to search for work that may be created by overly generous

unemployment benefits, a higher replacement rate may be socially optimal.13 Although the wage

replacement rate is about 50 percent for those receiving up to the maximum weekly benefit amount,

because that maximum has not kept pace with wage growth in most states, the share of UI benefit

recipients who receive half of lost wages has fallen. In addition, as of the start of 2020, the potential

duration of regular state UI benefits in 10 states had been reduced to fewer than 26 weeks, which had

been the common minimum standard since the 1950s. Because they reduce the value of the stream of

payments a claimant can expect to receive, lower real benefits and shorter potential durations can reduce

the incentive for an unemployed person to apply for benefits. Among those who do apply, the actual

average duration of benefit receipt is shorter; this reduces the average number of insured unemployed in

any week, which also lowers the recipiency rate.

12

Figure 1. Percent of Unemployed Receiving Regular State Unemployment Insurance Benefits, 1979–2019, Annual Averages

Source: U.S. Department of Labor, Employment and Training Administration, https://oui.doleta.gov/unemploy/data_summary/DataSum.asp

Figure 2. UI Recipiency Rate by State, Annual Average, 2019

Source: U.S. Department of Labor, Employment and Training Administration, https://oui.doleta.gov/unemploy/data_summary/DataSum.asp

13

These and related policies may reflect a change in attitude about the UI system among policymakers in

certain states. Instead of viewing unemployment insurance as a program to support productive job search

and improve the efficiency of job matches, the policy posture in some states is that unemployment

insurance is simply a business cost to be minimized. Low UI tax rates are a feature touted by economic

development business location marketers. To further reduce the tax burden of financing UI benefits, the

application process can be set up as a hurdle to benefit access. A particularly stark case is Florida, which

changed the process in 2011 to require, among other things, that applicants file through an automated

online system available only in English and take a 45-question online skills assessment test. Although

some of the new requirements were deemed discriminatory in legal challenges and ultimately changed,

the unemployment insurance filing process remained difficult, particularly for low-educated and low-

wage workers (Wentworth and McKenna, 2015). Across the country, benefit denials on procedural

grounds have increased significantly as states have tightened their filing and documentation requirements

(Wentworth, 2017).

The decline in the recipiency rate also is related to changes in the economy and the nature of work. For

example, the recipiency rate is positively associated with the rate of unionization (O’Leary and Wandner,

2020). One possible reason is that unions may provide valuable information about the UI application

process to members who are laid off. Declining rates of unionization may well have contributed to the

decline in the recipiency rate.

Additionally, growth in nonstandard work arrangements, along with the relative increase in low-wage

employment, likely has reduced the share of those losing their jobs who qualify for UI benefits. As noted

above, the use of independent contractors, who are not eligible for unemployment insurance, is

widespread in some industries; some research based on administrative data also suggests the size of the

independent contractor workforce has grown in recent years (Jackson, Looney, and Ramnath, 2017; Lim

et al., 2019). The use of temporary help and other third-party contracting arrangements has grown in

manufacturing, transportation and warehousing, and construction (Dey, Houseman, and Polivka, 2012,

2017; Bernhardt et al., 2016). In addition, the increased use of scheduling algorithms in retail, hospitality,

and other service industries has meant that workers in these sectors receive relatively little notice about

their work schedules, and their hours and earnings vary considerably from week to week (Henly and

Lambert, 2014; Lambert, Fugiel, and Henly, 2014; Lambert, Henly, and Kim, 2019).14 Both third-party

contracting arrangements, which often are associated with lower wages (Weil, 2014), and variable hours

and earnings may mean that employees do not meet minimum earnings requirements for unemployment

insurance.

14

The UI recipiency rate during the first quarter of 2020 jumped to 37 percent, according to U.S.

Department of Labor data, reflecting the surge in unemployment insurance claims in March that came

with the first wave of COVID-19 layoffs. With the number applying for regular UI benefits exceeding 1

million in all weeks in the second quarter of 2020, the second quarter recipiency rate should be at an

historic high. The very high numbers applying for and receiving unemployment insurance during the

pandemic reflect provisions included in the Coronavirus Aid, Relief, and Economic Security (CARES)

Act to increase the adequacy of benefits, thereby making it more attractive for the unemployed to apply

for benefits. As discussed above, the CARES Act also expanded coverage under the Pandemic

Unemployment Assistance program to many individuals, including the self-employed, who would not

have qualified for benefits under the regular state UI programs.

Recommendations: Expanding UI Coverage for Workers in Precarious and Nonstandard Arrangements

The economic fallout from the pandemic has focused policy attention on the inadequacies of the

unemployment insurance system in the United States. The provisions in the CARES Act pertaining to

unemployment insurance were temporary measures to deal with extraordinary circumstances. While a full

discussion of necessary UI reforms is beyond the scope of this essay, we discuss several steps that we

believe should be taken to expand benefits in more ordinary times to more job losers who do not qualify

for UI benefits and who often work in precarious or nonstandard work arrangements.15

ALLOW WORKERS TO COUNT THEIR MOST RECENT EARNINGS TOWARD DETERMINING ELIGIBILITY

Among other requirements, under the standard rules in place in most states, a worker must have earnings

in covered employment above a specified threshold during the first four of the five most recently

completed quarters to qualify for UI benefits. Eligibility for someone who became unemployed in mid-

June, for example, would be based on his or her earnings during the previous calendar year. Not counting

a person’s most recent earnings toward benefits qualification disadvantages those who are new to the

labor market or have irregular work histories. This rule may have made sense in the past when long data-

processing lags meant that earnings data for the most recent quarter typically were not available, but it

does not make sense today. As already noted, as of the start of 2019, 37 states and the District of

Columbia had adopted rules allowing workers who did not qualify for benefits using the standard

approach to use earnings during a more recent base period, typically the most recent four completed

15

quarters, to establish benefit eligibility (U.S. Department of Labor, 2020a). This should be the practice in

all states.

DETERMINE ELIGIBILITY BASED ON HOURS RATHER THAN EARNINGS

The primary rationale for requiring earnings above some threshold to qualify for unemployment benefits

is to ensure that recipients have a significant attachment to the labor market. Expressing the benefit

qualification threshold in terms of earnings rather than hours, however, significantly disadvantages low-

wage workers, who must work more hours to attain any given level of earnings. To help level the playing

field for low-wage workers, we recommend that other states adopt the approach already in place in the

state of Washington, where the requirement to qualify for unemployment benefits is a minimum number

of hours worked rather than a minimum level of earnings.

Few states currently collect the information from employers that would be needed to implement an hours

standard as the default approach to determining benefit eligibility. States should take steps to begin

collecting information on hours worked, but in the interim, there are alternative means of ensuring that

low-wage workers with significant labor market attachment qualify for benefits. First, states could adjust

their earnings qualification thresholds so that someone who worked, say, 20 hours per week at the state

minimum wage for at least one quarter during the base period would have enough earnings to qualify for

benefits. This is currently not the case in 28 states. Second, working more than some minimum number of

hours could be offered as an alternative route to qualifying for benefits for those who cannot satisfy the

minimum earnings requirement, with employers contacted to verify hours for workers who make claims

on that basis.

DROP THE REQUIREMENT THAT THE UNEMPLOYED SEEK FULL-TIME WORK

The common requirement that beneficiaries seek full-time work in order to qualify for UI benefits is

another factor that may limit access to benefits for precarious workers. Whether because of family

responsibilities or the nature of the jobs that are available to them, many precarious workers hold part-

time positions. If part-time work offers the best or most feasible employment option for an unemployed

individual, searching for such work should not disqualify that person from benefits. We recommend that

any unemployed worker searching for part-time work of 20 hours or more per week who otherwise

qualifies for unemployment insurance benefits be allowed to collect them.

16

REFORM PARTIAL UNEMPLOYMENT INSURANCE BENEFITS

Although every state provides for the payment of partial unemployment benefits under certain

circumstances, the benefit formulas are far from generous. In a majority of states, for example, a low-

wage worker whose earnings are cut in half would receive no benefits. The earnings reductions that

would be needed before any partial benefits would be payable is even larger for higher-wage workers. We

recommend that states reevaluate their partial unemployment benefit formulas to better protect workers

who lose a substantial fraction of their work hours, including in the case where this occurs because the

worker has lost a second job. Under recent changes made to partial benefits in Vermont, workers are

eligible for benefits up to the point where their earnings equal twice their weekly benefit amount and the

benefit payable is equal to the weekly benefit amount minus half the amount the worker has earned. These

provisions accomplish exactly what we have in mind. We recommend that other states raise the maximum

amount their workers can earn without losing eligibility for partial benefits and adopt benefit formulas

thatreduce the benefit amount by some fraction of the amount earned up to that maximum earnings level

rather than dollar for dollar.

SIMPLIFY THE APPLICATION PROCESS

Several other reforms to the UI system could help precarious workers with unstable employment and

irregular earnings to access benefits. One simple yet especially important change would be to make it

easier to apply for unemployment benefits. The COVID-19 crisis has drawn attention to the lack of

ongoing investment in the IT infrastructure needed for an unemployment insurance system to operate

smoothly. Modernizing this infrastructure should be a priority and would do a great deal to ensure that

states are able to process benefit applications efficiently and expeditiously.

State policy choices also matter. Many states have moved to having workers file for benefits primarily or

even exclusively online, and reductions in staffing at many state unemployment agencies mean that

reaching someone on the telephone or in person can be difficult (Balducchi and O’Leary, 2018). Some

states also have imposed additional requirements, such as that workers collecting unemployment benefits

post a résumé online through the employment service, another task that a less-educated worker may find

difficult. While there is a legitimate interest in ensuring that those collecting unemployment insurance

benefits are, in fact, entitled to them, states should make the process of applying for benefits as

straightforward as possible and, in particular, should make telephone or in-person application and

reporting options readily available to those who need them. Staff assistance also should be available to

17

any UI claimant undertaking an active job search to provide help with résumé preparation and job

applications.

PROVIDE A SAFETY NET FOR SELF-EMPLOYED WORKERS

A significant challenge in the modern economy is how to provide a safety net for self-employed workers

who generally are ineligible for unemployment insurance. It is difficult to know whether a self-employed

person is working less because no work is available or by choice. For this reason, although this has been

done in some European countries, we do not advocate for blanket extension of coverage to the self-

employed under the existing unemployment insurance program. On the other hand, as discussed, the self-

employed are a heterogeneous group that includes business owners, independent contractors (including

gig workers who obtain work through online platforms or mobile apps), and workers in informal

nonemployee arrangements. Changes in technology and other factors are likely to contribute to growing

rates of self-employment in the coming years, and many of these workers are likely to be exposed to the

risk of considerable income fluctuation.

There are examples of unemployment insurance coverage being extended to the self-employed on a

voluntary basis. In California, for example, business owners who have employees may elect to be covered

under the state’s unemployment insurance program. As already discussed, there are also some European

countries in which the self-employed are allowed to opt into unemployment insurance coverage. In

general, however, few self-employed people choose to be covered when enrollment is voluntary.

Another approach sometimes proposed for providing unemployment insurance coverage to the self-

employed involves establishing self-funded individual security accounts (see, for example, Kletzer and

Rosen, 2006; Feldstein and Altman, 2007; Gruber, 2016). Some proposals would establish such accounts

for everyone, but our main interest lies with how they might function for the self-employed. The basic

idea is that self-employed workers would pay a small fraction of their wages into individual funds that

they could draw upon in the event of not being able to find work (or, in some cases, if other needs arose).

Some proposals cap the amount that can be accumuled, with money not spent before retirement rolled

over into a retirement account. The purported advantage of these accounts is that they would ameliorate

the moral hazard problem associated with conventional unemployment insurance, because money taken

from the account would in effect come out of the worker’s own pocket. Even assuming that workers are

sufficiently forward-looking to understand that taking money out of the account today could leave them

less well-off in the future, a problem arises when the account balance falls to zero. If workers can only

18

withdraw up to the amount they have contributed, such accounts might fail to protect those who need

unemployment insurance the most. Yet, if workers are entitled to draw benefits even when their account

balance is negative, the moral hazard problem is reintroduced.

Instead, we advocate for two other approaches we believe are more viable. The first pertains to

independent contractors, many of whom rely primarily or exclusively on a single client for work.

Research has concluded that a sizable number of these workers should have been categorized as

employees rather than as independent contractors (see, e.g., General Accountability Office, 2009; Carré,

2015; National Employment Law Project, 2017). Workers who are misclassified do not have access to

unemployment insurance or workers’ compensation. Additionally, worker misclassification is strongly

associated with wage and hours violations, including wage theft, and with underpayment of taxes (Weil,

2014, 2017; Carré, 2015). Employers may realize substantial cost savings by classifying their workers as

independent contractors, which, coupled with weak enforcement of existing laws, incentivizes employers

to misclassify their workers (Belman and Sojourner, 2019).

To address misclassification, we urge greater clarity in who may be classified as an independent

contractor. Rules governing classification should be based on the fundamental realities of the employment

relationship and the locus of control, and they should discourage legal maneuvers intended to shield

companies from misclassification lawsuits—such as the formation of long subcontracting chains that

artificially separate employers from workers and the establishment of what many argue are sham limited

liability companies. Moreover, existing rules must be enforced. A 2009 report from the Government

Accountability Office called for greater coordination and enforcement to reduce worker misclassification,

and in the initial years following the report, considerable progress was made in improving collaboration

across federal agencies and states to tackle the problem (Weil, 2017). Those initiatives should be

restarted. Such actions could significantly reduce the ranks of the self-employed and yield many benefits

for workers, including access to unemployment insurance.

Of course, not all self-employed workers should be reclassified as employees. A complementary policy

therefore would provide protection to the most vulnerable by establishing a means-tested Job Search

Allowance intended to assist self-employed individuals who are seeking jobs. As we envision this

allowance, it would be available to self-employed individuals with earnings histories comparable to those

required to qualify for normal unemployment insurance benefits. It would pay a flat weekly benefit

amount of half the average weekly benefit paid to unemployed workers in the state. The benefit would be

funded out of general revenues. As with unemployment insurance, beneficiaries would be required to be

19

actively engaged in searching for jobs. Such a benefit would help self-employed individuals whose

business activity has failed while they transition into new work.

PROVIDE UNEMPLOYMENT INSURANCE AS A STIPEND DURING TRAINING

If increasing numbers of workers lose their jobs because of new technology adoption, globalization, or

other economic disruptions, more will require training to find new work. Current federal funding for

training of displaced workers is limited. In program year 2018, only 11.2 percent of dislocated workers

participating in programs funded by the federal Workforce Innovation and Opportunity Act (WIOA)

received any training (U.S. Department of Labor, 2020b, Table III-12). Other training dollars are

available from various state and private funds, but arguably these amounts are inadequate and would need

to be significantly increased to meet demand, even if the rate of worker displacement from technology

does not rise (Hanks and Madland, 2018).

The unemployment insurance system can play an integral role in identifying workers in need of training

and providing a stipend during the training period. The federal-state workforce system, largely operates

through a network of approximately 2,400 American Job Centers located in communities across the

country. Over the last couple of decades, great strides have been made to better integrate and coordinate

service delivery, with the colocation of various public employment and training services in the job

centers, also known as “one-stops.” In some areas, unemployment insurance services also are provided

alongside services in the job centers. Nevertheless, program silos still exist. Job center staff, for example,

do not routinely inform clients about other training dollars that might be available through state and

privately funded programs. Using the job centers as clearinghouses for information on funding for job

training could significantly improve service delivery to those in need (Hershbein, 2015).

In addition to needing money to cover the costs of training, workers who lose their jobs and undergo job

training may need income support while they train for new jobs. This is especially likely for workers from

low-income households. For many, unemployment insurance can fill that need. The work-search

requirement may be waived for those participating in training through WIOA, and job center staff also

may be able to waive the work search requirement for those who participate in training through other

programs. We anticipate that the UI system could play an even more important role in the future in

providing stipends for displaced workers who require training.

20

Conclusion

Although technological advances and globalization carry substantial benefits, they also harm many who

lose their jobs and livelihoods as a result. A strong social safety net is essential for supporting those

harmed by change and, more broadly, for garnering widespread public acceptance of change. The

unemployment insurance system is the major vehicle for providing income assistance to those who lose

their jobs through no fault of their own. Particularly in recent years, however, a declining share of the

unemployed has received unemployment insurance, in part because fewer of those who lose their jobs

qualify for benefits. And, many who are left out of the unemployment insurance system are particularly

vulnerable to economic change and dislocation.

This situation has arisen because the basic structure of the unemployment insurance system in the United

States has not adequately adapted to changes in the nature and organization of work. Goals of much-

needed reform to the UI system should include better protecting those in precarious and nonstandard work

arrangements by expanding coverage to low-wage and part-time workers and those with unstable work

schedules. Providing greater access to benefits among employees in precarious arrangements will cost

money—as will reforms to address other pressing UI issues such as declining wage replacement rates—

and this means their implementation will require changes to the existing UI financing structure. UI

benefits are financed through a payroll tax on employee wages. Federal law stipulates that, at a minimum,

states must tax the first $7,000 of an employee’s earnings—an amount that has not increased since 1983.

Many states have higher taxable wage bases, but the amount of wages that are taxed generally has not

kept pace with inflation (von Wachter, 2019). Although a full discussion of necessary UI reforms is

beyond the scope of this essay, one important component will be to strengthen UI financing by expanding

the taxable wage base and indexing it to inflation (O’Leary and Wandener, 2018).

Problems of moral hazard make extending unemployment insurance to the self-employed particularly

challenging. Nevertheless, the swelling ranks of independent contractors and gig workers and the

economic vulnerability of these groups also warrant a policy response. Clarifying and enforcing

regulations on who may be treated as an independent contractor will increase the number who are

classified as employees and are thereby covered by unemployment insurance and other social insurance

programs. For those who are self-employed and lose their jobs, providing a modest, means-tested, and

time-limited job search assistance stipend, funded through general revenues, would be an effective way to

provide income assistance to those most in need.

21

References

Abraham, Katharine G., John C. Haltiwanger, Kristin Sandusky, and James R. Spletzer. In press. “Measuring the Gig Economy: Current Knowledge and Open Issues.” In C. Corrado, J. Haskel, J. Miranda, and D. Sichel, eds., Measuring and Accounting for Innovation in the 21st Century. Chicago: University of Chicago Press.

Abraham, Katharine G., John C. Haltiwanger, Claire Hou, Kristin Sandusky, and James R. Spletzer. Forthcoming. “Reconciling Survey and Administrative Measures of Self-Employment.” Journal of Labor Economics.

Abraham, Katharine G., Brad Hershbein, and Susan N. Houseman. 2019. “Independent Contract, Informal, and Online Intermediary Work: Preliminary Evidence on Developing Better Measures in Household Surveys.” Unpublished working paper. Available: http://conference.iza.org/conference_files/Statistic_2019/abraham_k16798.pdf.

Abraham, Katharine G., and Susan N. Houseman. Forthcoming. Contingent and Alternative Employment: Lessons from the Contingent Worker Supplement, 1995–2017. Report to the U.S. Department of Labor.

Acemoglu, Daron, David Autor, David Dorn, Gordon H. Hanson, and Brendan Price. 2016. “Import Competition and the Great U.S. Employment Sag of the 2000s.” Journal of Labor Economics 34(S1): S141-S198.

Armstrong, Ken, Justin Elliott, and Ariana Tobin. 2020. “Meet the Customer Service Reps for Disney and Airbnb Who Have to Pay to Talk to You.” ProPublica. Available: https://www.propublica.org/article/meet-the-customer-service-reps-for-disney-and-airbnb-who-have-to-pay-to-talk-to-you. Accessed October 7, 2020.

Autor, David H., David Dorn, and Gordon H. Hanson. 2013. “The China Syndrome: Local Labor Market Effects of Import Competition in the United States.” American Economic Review 103(6): 2121–2168.

Autor, David, and Elisabeth Reynolds. 2020. “The Nature of Work after the COVID Crisis: Too Few Low-Wage Jobs.” The Hamilton Project Essay 2020-14. July.

Balducchi, David E., and Christopher J. O’Leary. 2018. “The Employment Service–Unemployment Insurance Partnership: Origin, Evolution, and Revitalization.” In S.A. Wandner, ed., Unemployment Insurance Reform: Fixing a Broken System, 65-102. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.

Belman, Dale L., and Richard Block. 2009. “Informing the Debate: The Social and Economic Costs of Misclassification in Michigan.” Institute for Public Policy and Social Research, Michigan State University. Available: http://ippsr.msu.edu/publications/ARMisClass.pdf. Accessed October 7, 2020.

Belman, Dale, and Aaron Sojourner. 2019. “Economic Analysis of Incentives to Fraudulently Misclassify Employees in District of Columbia Construction.” In Illegal Worker Misclassification: Payroll Fraud in the District’s Construction Industry. Issue Brief and Economic Report. Office of the Attorney General, District of Columbia. Available: https://oag.dc.gov/sites/default/files/2019-09/OAG-Illegal-Worker-Misclassification-Report.pdf. Accessed October 11, 2020.

22

Bernhardt, Annette, Rosemary Batt, Susan N. Houseman, and Eileen Appelbaum. 2016. “Domestic Outsourcing in the United States: A Research Agenda to Assess Trends and Effects on Job Quality.” Upjohn Institute Working Paper No. 16-253.

Bernhardt, Annette, Ruth Milkman, Nik Theodore, Douglas D. Heckathorn, Mirabaj Auer, James deFilippis, Ana Luz González, Victor Narro, Jason Perelshteyn, Diana Polson, and Michael Spiller. 2009. Broken Laws, Unprotected Workers: Violations of Employment and Labor Laws in American Cities. Washington, DC: National Employment Law Project.

Blaustein, Saul A. 1993. Unemployment Insurance in the United States: The First Half Century. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research.

Brynjolfsson, Erik, and Andrew McAfee. 2014. The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies. New York, NY: W.W. Norton and Co.

Bureau of Labor Statistics. 2018. Electronically Mediated Work: New Questions in the Contingent Worker Supplement. Monthly Labor Review. September. Available: https://www.bls.gov/opub/mlr/2018/article/pdf/electronically-mediated-work-new-questions-in-the-contingent-worker-supplement.pdf. Accessed October 7, 2020.

Carré, Francoise. 2015. (In)dependent Contractor Misclassification. Washington, DC: Economic Policy Institute.

Casselman, Ben, Patricia Cohen, Conor Dougherty, and Nelson D. Schwartz. 2020. “A Lifeline to the Jobless Has Problems with Fraud, and with Math.” The New York Times, September 11, 2020.

de Silva, Lalith, Adrian W. Millett, Dominic M. Rotondi, and William F. Sullivan. 2000. Independent Contractors: Prevalence and Implications for Unemployment Insurance Programs. Planmatics, Inc., Report to U.S. Department of Labor Employment and Training Administration.

Dey, Matthew, Susan Houseman, and Anne Polivka. 2012. “Manufacturers’ Outsourcing to Staffing Services.” Industrial and Labor Relations Review 65(3): 533–559.

Dey, Matthew, Susan Houseman, and Anne Polivka. 2017. “Manufacturers’ Outsourcing to Temporary Help Services: A Research Update.” Bureau of Labor Statistics Working Paper 493.

Dillender, Marcus, Carolyn Heinrich, and Susan Houseman. Forthcoming. “Effects of the Affordable Care Act on Part-Time Employment.” Journal of Human Resources.

Feldstein, Martin S. 1974. "Unemployment Compensation: Adverse Incentives and Distributional Anomalies," National Tax Journal 27 (June): 231-244.

Feldstein, Martin, and Daniel Altman. 2007. “Unemployment Insurance Savings Accounts.” In J. Poterba, ed., Tax Policy and the Economy, Volume 21, 35-63. Cambridge, MA: MIT Press.

General Accountability Office. 2009. Employee Misclassification: Improved Coordination, Outreach, and Targeting Could Better Ensure Detection and Prevention. Washington, DC.

23

Gruber, Jonathan. 2016. Security Accounts as Short Term Social Insurance and Long Term Savings: Expanding Financial Security for Workers in the New Economy. Fresh Perspective Series, Aspen Institute Future of Work Initiative.

Hanks, Angela, and David Madland. 2018. Better Training and Better Jobs: A New Partnership for Sectoral Training. Center for American Progress.

Harris, Seth D., and Alan B. Krueger. 2015. “A Proposal for Modernizing Labor Laws for Twenty-First-Century Work: The ‘Independent Worker.’” The Hamilton Project Discussion Paper 2015-10.

Henly, Julia R., and Susan J. Lambert. 2014. “Unpredictable Work Timing in Retail Jobs: Implications for Employee Work-life Outcomes.” Industrial and Labor Relations Review 67(3): 986–1016.

Hershbein, Brad. 2015. “Federal Education Grant Money Is a Terrible Thing to Waste.” Brookings Institution. Available: https://www.brookings.edu/blog/social-mobility-memos/2015/09/21/federal-education-grant-money-is-a-terrible-thing-to-waste/amp/. Accessed October 1, 2020.

Jackson, Emilie, Adam Looney, and Shanthi Ramnath. 2017. “The Rise of Alternative Work Arrangements: Evidence and Implications for Tax Filing and Benefit Coverage.” Office of Tax Analysis Working Paper 114. January.

Katz, Lawrence F., and Alan B. Krueger. 2019. “The Rise and Nature of Alternative Work Arrangements in the United States, 1995-2015.” ILR Review (March) 72(2): 382–416.

Kletzer, Lori G., and Howard Rosen. 2006. “Reforming Unemployment Insurance for the Twenty-First Century.” The Hamilton Project Discussion Paper 2006-06. September.

Lambert, S.J., P.J. Fugiel, and J.R. Henly. 2014. Schedule Unpredictability among Early Career Workers in the U.S. Labor Market: A National Snapshot. EINet (Employment Instability, Family Well-being, and Social Policy Network) at the University of Chicago. Available: https://ssa.uchicago.edu/sites/default/files/uploads/lambert.fugiel.henly_.executive_summary.b_0.pdf. Accessed October 1, 2020.

Lambert, Susan J., Julia R. Henly, and Jaeseung Kim. 2019. “Precarious Work Schedules as a Source of Economic Insecurity and Institutional Distrust.” RSF: The Russell Sage Foundation Journal of the Social Sciences, 5(4): 218–257.

Lim, Katherine, Alicia Miller, Max Risch, and Eleanor Wilking. 2019. “Independent Contractors in the U.S.: New Trends from 15 Years of Administrative Tax Data.” Unpublished working paper. Available: https://www.irs.gov/pub/irs-soi/19rpindcontractorinus.pdf. Accessed October 1, 2020.

Maquet, Isabelle, Virginia Maestri, and Céline Thévenot. 2016. “The Coverage Rate of Income Support Measures in the EU: Measurement and Challenges.” European Commission Working Paper 2/2016.

Matsaganis, Manos, Erhan Özdemir, Terry Ward, and Alkistis Zavakou. 2015. “Non-standard Employment and Access to Social Security Benefits.” European Commission Research Note 8/2015.

National Commission on Unemployment Compensation (NCUC). 1980. National Commission on Unemployment Compensation: Final Report. Washington, D.C. July.

24

National Employment Law Project. 2017. Independent Contractor Misclassification Imposes Huge Costs on Workers and Federal and State Treasuries. Washington, DC: National Employment Law Project.

OECD. 2018a. Employment Outlook 2018. Paris: OECD Publishing.

OECD. 2018b. The Future of Social Protection: What Works for Non-standard Workers? Paris: OECD Publishing.

OECD. 2020 Benefits in unemployment, share of previous income (indicator). Available: https://www.oecd-ilibrary.org/employment/benefits-in-unemployment-share-of-previous-income/indicator/english_0cc0d0e5-en. Accessed November 8, 2020.

O'Leary, Christopher J. 2011. "Benefit Payment Costs of Unemployment Insurance Modernization: Estimates Based on Kentucky Administrative Data." Upjohn Institute Working Paper 11-172.

O’Leary, Christopher J., and Kenneth J. Kline. 2020. “State Unemployment Insurance Reserves Are Not Adequate.” Upjohn Institute Working Paper 20-321.

O’Leary, Christopher J., and Robert A. Straits. 2004. "Intergovernmental Relations in Employment Policy: The United States Experience." In A. Noël, ed., Federalism and Labour Market Policy: Comparing Different Governance and Employment Strategies, 25-82. Montreal and Kingston: McGill-Queen's University Press.

O’Leary, Christopher J., and Stephen A. Wandner. 2018. “Unemployment Insurance Reform: Evidence-Based Policy Recommendations,” Chapter 5 (pp. 131-210) in: Unemployment Insurance Reform: Fixing a Broken System, Stephen A. Wandner, editor. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research, 2018.

O’Leary, Chris, and Stephen A. Wandner. 2020. “An Illustrated Case for Unemployment Insurance Reform.” Upjohn Institute Working Paper 19-317.

Penn, Ben. 2020. “DOL Debuts Rule Easing Business Use of Independent Contractors.” Daily Labor Report. September 22.

Perea, Juan F. 2011. “The Echoes of Slavery: Recognizing the Racist Origins of the Agricultural and Domestic Worker Exclusion from the National Labor Relations Act.” Ohio Stat Law Journal, 72(1): 95-138.

Pierce, Justin, and Peter Schott. 2016. “The Surprisingly Swift Decline of U.S. Manufacturing Employment.” American Economic Review 106(7): 1632–1662.

Rosenberg, Eli. 2020. “New Trump Administration Rule Could Make It Harder for Gig and Contract Workers to Have Rights as Employees.” The Washington Post. September 22, 2020.

U.S. Advisory Council on Unemployment Compensation (USACUC). 1996. "Collected Findings and Recommendations, 1994-1996." Washington, D.C.: Advisory Council on Unemployment Compensation.

25

U.S. Department of Labor, Employment, and Training Administration. 2019. Comparison of State Unemployment Insurance Laws. Available: https://oui.doleta.gov/unemploy/pdf/uilawcompar/2019/complete.pdf. Accessed July 25, 2020.

U.S. Department of Labor, Employment and Training Administration. 2020a. Significant Provisions of State Unemployment Insurance Laws Effective January 2020. Available:https://oui.doleta.gov/unemploy/content/sigpros/2020-2029/January2020.pdf. Accessed August 23, 2020.

U.S. Department of Labor, Employment and Training Administration. 2020b. PY 2018 Data Book. February 2020b. Available: https://www.doleta.gov//performance/results/WIASRD/PY2018/PY_2018_WIOA_and_Wagner-Peyser_Data_Book.pdf. Accessed August 28, 2020.

Valletta, Robert G., Leila Bengali, and Catherine van der List. Forthcoming. “Cyclical and Market Determinants of Involuntary Part-Time Employment.” Journal of Labor Economics.

Von Wachter, Till. 2019. “Unemployment Insurance Reform.” Annals of the American Academy of Political and Social Science, 686: 121-146.

Vroman, Wayne. 2009. “Unemployment Insurance Recipients and Nonrecipients in the CPS.” Monthly Labor Review 132 (October): 44-53. Washington, DC: U.S. Department of Labor, Bureau of Labor Statistics.

Weil, David. 2014. The Fissured Workplace. Cambridge, MA: Harvard University Press.

Weil, David. 2017. “Lots of Employees Get Misclassified as Contractors. Here’s Why It Matters. Harvard Business Review. July 5.

Wentworth, George. 2017. Closing Doors on the Unemployed: Why Most Jobless Workers Are Not Receiving Unemployment Insurance and What States Can Do About It. Washington, DC: National Employment Law Project.

Wentworth, George, and Claire McKenna. 2015. “Ain’t No Sunshine: Fewer than One in Eight Unemployed Workers in Florida Is Receiving Unemployment Insurance.” National Employment Law Project Policy Brief. September 21.

West, Rachel, Indivar Dutta-Gupta, Kali Grant, Melissa Boteach, Claire McKenna, and Judy Conti. 2016. Strengthening Unemployment Protections in America, Center for American Progress, National Employment Law Project, and Georgetown University Center on Poverty and Inequality.

26

Endnotes

1 The replacement rates calculated by the OECD compare net income in unemployment to net income in employment. In addition to unemployment benefits, the net replacement rates reflect other cash benefits that may be available to unemployed individuals, such as means-tested unemployment compensation, housing benefits or family assistance. These other benefits often differ by family type, with many countries providing additional assistance to families with children.

2 The 10 states that reduced potential durations to less than 26 weeks following the Great Recession were Arkansas, Florida, Georgia, Idaho, Illinois, Kansas, Michigan, Missouri, North Carolina, and South Carolina. Illinois later restored the maximum potential duration to 26 weeks, but as of the start of 2020, the maximum duration in Alabama had been reduced to 14 weeks. In some of these states, the maximum duration of benefits is tied to the state unemployment rate (O’Leary and Kline 2020, U.S. Department of Labor 2020a).

3 Today, agricultural employers who pay $20,000 or more in wages in a quarter or employ more than 10 workers, as well as households that pay $1,000 or more in wages to domestic employees in a quarter are required to provide UI coverage. Religious organizations are not required to cover employees who have duties of a religious nature, but in many states they are required to cover employees performing secular duties.

4 As discussed by Casselman et al. (2020), the long lags that have been experienced in processing PUA claims have complicated the interpretation of Department of Labor data on the number of claims. At the point when a person’s claim is approved, the data reflect all of the weeks for which retroactive payment is being made. In addition, some states have experienced problems with fraud. Although fraudulent claims may be weeded out, unless or until that happens, they are reflected in the claim counts.

5 In the state of Washington, eligibility for benefits is based on hours worked rather than earnings. In the remaining 21 states and the District of Columbia, working 20 hours per week for three months at the state’s minimum wage would satisfy the base period earnings requirement.

6 The CPS supplement on UI filers and nonfilers has been administered five times (1976, 1989, 1993, 2005, and 2018).

7 These figures are based on the authors’ tabulations of the data. Vroman (2009) provides a detailed analysis of the 2005 nonfilers supplement.

8 Workers may access pro-rated benefits if their employer sets up a work share or short-time compensation (STC) plan with the state. Currently, 26 states offer an STC program as part of the state’s unemployment insurance system. Under STC, an employer faced with a temporary reduction in business reduces workers’ hours in lieu of laying off workers. The work share plan allows employees to collect pro-rated unemployment benefits.

9 Recognizing this situation, Harris and Krueger (2015) proposed creating a third category of worker that they termed independent workers. This category, which would apply to independent contractors such as ride-share workers deemed to occupy a gray area, would be granted some worker protections not given others who are self-employed.

10 Workers who obtained work through online platforms or mobile apps that mediated the payment for their services made up only about 1 percent of the employed in 2017, according to data collected in the Bureau of Labor Statistics Contingent Worker Supplement (Bureau of Labor Statistics, 2018).

11 Carré (2015) notes that because the purpose of audits is to identify fraud and recover tax revenues, they do not randomly sample businesses. Researchers extrapolate from the audit studies to estimate the national incidence. The direction of any bias in these estimates, however, is unclear. The National Employment Law Project (2017) argues that studies based on audit data may understate the incidence of misclassification because misclassification is especially likely among employers who do not report any workers as employees, and these employers are not audited.

12 During recessions, the duration of benefits often is extended under programs funded partially or entirely by the federal government. Our focus here, however, is on the regular state UI programs.

13 For a summary of this research, see von Wachter (2019). 14 There has been a notable structural increase in involuntary part-time employment in service industries,

particularly retail and hospitality, which make heavy use of scheduling algorithms (Dillender, Heinrich, and Houseman, forthcoming; Valletta, Bengali, and van der List, forthcoming).

27

15 Other critical issues for UI reform include improving benefit adequacy and putting the system financing on a

sounder footing.